Research On Tracking Method Of Moving Targets In Complex Environment

Radar can locate,track and recognize the targets by utilizing electromagnetic wave,which is widely used in military and civilian fields.And radar target tracking plays an extremely important role in modern radar systems.With the development and progress of the science technology and society,the environment of the targets is becoming more and more complex,and the maneuverability of the targets keeps enhancing,which leads to higher requirements for tracking methods of radar target.As a result,it is of great theoretical research and practical application value to carry out research on the tracking methods of moving targets in complex environment.The main contents of this thesis are as follows.First of all,this thesis focuses on the analysis and comparison of several target tracking algorithms under noise background.The Kalman filtering algorithm performs well in non-maneuvering target tracking.In addition,several algorithms can obtain satisfying processing results in maneuvering target tracking,which include Singer model algorithm,the current statistical model algorithm,and the interacting multiple model algorithm.Simulation results show the following facts.Compared with the Singer model algorithm,the tracking accuracy of the current statistical model algorithm is higher under general maneuvering and high maneuvering condition.But the above two algorithms are similar in tracking accuracy under low maneuvering condition.On the whole,the tracking accuracy of the interacting multiple model algorithm is higher than that of the other two algorithms under the three above conditions.Next,in view of the problem that the computation of joint probabilistic data association algorithm will increase rapidly when the number of targets or the clutter density increases,a method based on m-best joint probabilistic data association algorithm is introduced.And the theoretical analysis and simulation experiments of several classical target tracking algorithms in clutter environment are given.As the results indicate,both the nearest neighbor algorithm and the probabilistic data association algorithm can track single target in clutter environment.However,when the clutter density is excessively high,the nearest neighbor algorithm may lose targets.Compared with the single target tracking problem in clutter environment,the multiple targets tracking problem in clutter environment needs to comprehensively consider the source of the candidate echoes located in the intersecting gates.The joint probabilistic data association algorithm confirms the source of candidate echoes by splitting the confirmation matrix,so as to track multiple targets in clutter environment.But the splitting of the confirmation matrix is a combinatorial number problem.With the increase of the figure for the targets or the enlargement of the clutter density,the computational complexity of the joint probabilistic data association algorithm will increase rapidly.Therefore,a method based on m-best joint probabilistic data association algorithm is introduced.This algorithm transforms the processing of searching joint events into a linear programming problem.Moreover,this algorithm solves the first m optimal solutions of the linear programming iteratively,which significantly reduces the processing time of the joint probabilistic data association algorithm.Finally,an improved interacting multiple model joint probabilistic data association algorithm is proposed to track multiple maneuvering targets in clutter environment.First of all,the interacting multiple model joint probabilistic data association algorithm is implemented,which is the combination of the interacting multiple model algorithm for maneuvering target tracking and the joint probabilistic data association algorithm for multiple targets tracking.Then,with the increase of the figure for the targets or the enlargement of the clutter density,the computation of the interacting multiple model joint probabilistic data association algorithm will increase rapidly.To solve this problem,an improved interacting multiple model joint probabilistic data association algorithm is proposed,which is based on the unified gates to select candidate echoes and the m-best joint probabilistic data association algorithm.Simulation results show that the interacting multiple model joint probabilistic data association algorithm and the improved interacting multiple model joint probabilistic data association algorithm perform well for multiple maneuvering targets tracking in clutter environment.However,with the increase of the figure for the targets,the improved interacting multiple model joint probabilistic data association algorithm this thesis proposed has a smaller computation and a slightly higher tracking accuracy.